Between 2013 and 2014 two of the most prestigious scientific journals, Science and Cell, published a pair of studies from the same lab reporting that the protein Growth Differentiation Factor 11 (GDF11) was able to bring back youth to older heart and muscle cells. This discovery was highlighted by Science as one of the top scientific breakthroughs in 2014. It caused a deep impact on public opinion (for articles about “the secret to eternal youth” check here and here) and also on the scientific community, which quickly launched studies and clinical trials to test the potential of GDF11 in other areas of regenerative medicine such as Alzheimer. Sadly this is as far as the fairy tale goes. A study has been published in the last issue of Cell Metabolism (another very prestigious journal) repeating the experiments and disproving most of the claims.
The original studies were based on the seminal discovery that when older mice were exposed to younger blood their heart (2013) and muscle cells (2014) recovered some features of their youth. Experiments were conducted to identify the “rejuvenating factor” by comparing younger versus older blood. The results showed significantly different levels of one protein: GDF11. Next, it was confirmed that GDF11’s levels decrease with age. And importantly, when GDF11 was given to older mice, their heart and muscle cells looked younger and regenerated better. This is a logical sequence of scientific experiments to show that GDF11 plays a key role in ageing and could be a promising therapy for degenerative diseases.
However, some proteins of the same family as GDF11 are known to prevent rather than improve muscle regeneration, and a second group decided to take a closer look. Of note, they focused on the role of GDF11 in the story (they did not address the effect of young blood on older tissues). They found that GDF11 does not decrease with age. It actually increases. And they explain these discrepancies by convincingly showing that the original paper was flawed by the use of imprecise reagents in several steps. Furthermore, the second group reported that GDF11 treatment inhibits rather than promotes muscle regeneration. This last discrepancy is even more worrying because it is difficult to blame technical problems for the results. The more likely explanations are over-interpretation, misinterpretation or fabrication of the data. And hence the scandal.
Science is a self-correcting process. It relies on repetition and replication of results, so if something is false, it should be picked up. By in large this is an effective and efficient process which allows confirmation and building upon previous work. However, we live in the “publish or perish” era: a scientist’s publication record alone can determine whether they get funding to pay their salaries and continue their projects. This puts researchers under great pressure to publish their work in high impact journals.
This system has two major flaws: firstly, it leads to the publication of immature or fabricated data with exaggerated or wrong conclusions. Secondly, it discourages scientists from repeating the work of others, as replicated work can rarely be published in a high impact journal. Due to this lack of incentive to repeat studies, we currently have very little sense of the number of articles containing mistaken or false data, like the GDF11 one. If there are a large number, these studies could be polluting our understanding of a topic, resulting in a huge amount of wasted money and time. Actually, this week PLOS Biology has published a report estimating that irreproducible biology research costs $28 billion per year (Liam will discuss this issue in a blog post later this week).
The scientific community has been grappling with this issue over the last few years. It is necessary to find a measure of the output of scientists (especially when they are funded from the public purse), but the system should not compromise the quality of the data. Alternative publishing methods have been suggested, but, for the time being, the status quo remains unchanged.